voculinux/CONTRIBUTING.md

# Contributing to Voculinux

Thanks for your interest in contributing. This guide covers setup, style, testing, and pull requests.

## Code of conduct

Participation is covered by our [Code of Conduct](CODE_OF_CONDUCT.md). Be respectful and constructive.

## Ways to contribute

- **Report bugs** - [Open an issue](https://git.bongbetic.com/xavierk/voculinux/issues/new)
- **Suggest features** - [Open an issue](https://git.bongbetic.com/xavierk/voculinux/issues/new) with the capability you want
- **Improve documentation** - Fixes and clarity are always useful
- **Fix bugs or add features** - Check [open issues](https://git.bongbetic.com/xavierk/voculinux/issues)

## Development setup

### Automated (recommended)

```bash
git clone https://git.bongbetic.com/YOUR-USERNAME/voculinux.git
cd voculinux
./install.sh --dev
```

This installs system dependencies, creates `venv/` with an editable install and dev
tools, and runs the test suite. See [The two environments](#the-two-environments).

### Manual setup

1. Fork and clone the repository.

2. Install system dependencies (examples). Voculinux is published only as a
   Void Linux xbps package; the distro examples below are for building from a
   source checkout:

   **Ubuntu 24.04+:**
   ```bash
   sudo apt update
   sudo apt install -y python3-pip python3-gi python3-gi-cairo \
        gir1.2-gtk-3.0 gir1.2-ibus-1.0 libgirepository-2.0-dev libgirepository1.0-dev \
        build-essential pkg-config libcairo2-dev libffi-dev \
       python3-dev portaudio19-dev python3-venv xdotool
   ```

   **Debian 12 (installer environment only):**
   ```bash
   sudo apt install -y python3-pip python3-gi python3-gi-cairo \
        gir1.2-gtk-3.0 gir1.2-ibus-1.0 libgirepository1.0-dev libcairo2-dev \
        build-essential pkg-config libffi-dev \
       python3-dev portaudio19-dev python3-venv xdotool
   ```

   **Debian 13+:**
   ```bash
   sudo apt install -y python3-pip python3-gi python3-gi-cairo \
        gir1.2-gtk-3.0 gir1.2-ibus-1.0 libgirepository-2.0-dev libcairo2-dev \
        build-essential pkg-config libffi-dev \
       python3-dev portaudio19-dev python3-venv xdotool
   ```

   **Fedora 40+:**
   ```bash
   sudo dnf install -y python3-pip python3-gobject gtk3 ibus-devel \
        gobject-introspection-devel glib2-devel python3-devel portaudio-devel \
        cairo-devel cairo-gobject-devel libffi-devel python3-virtualenv pkg-config cmake \
        gcc-c++ make xdotool
   ```

   AppIndicator (system tray):
   ```bash
   # Older Ubuntu:
   sudo apt install -y gir1.2-appindicator3-0.1
   # Debian 12+ or newer Ubuntu:
   sudo apt install -y gir1.2-ayatanaappindicator3-0.1
   ```

   **macOS (tooling only):** `brew install portaudio pkg-config`; use
   `xcode-select --install` if a C compiler is missing. PyGObject/pycairo and their
   header checks apply only on Linux, which is required to run the desktop app.

3. Create the environment and install:

   ```bash
   # Requires uv (https://docs.astral.sh/uv/); creates .venv/ with dev + vad extras
   just deps
   ```

   Missing native build headers are reported with install commands before syncing.
   For older Linux releases, use the installer environment described below.

4. Run:

   ```bash
   just run-source-debug
   ```

5. Optional pre-commit hooks:

   ```bash
   uv run --no-sync pre-commit install
   ```

   Hooks are optional. CI runs the same checks.

### The two environments

The repository uses two virtual environments on purpose. Do not merge them:

| Directory | Created by | Python | Used for |
|-----------|-----------|--------|----------|
| `.venv/`  | `just deps` (uv) | pinned in `.python-version` | dev tooling: pytest, black, mypy, `just` recipes |
| `venv/`   | `./install.sh` | the system Python | running the installed app, which needs distro PyGObject |

- On Linux, `.venv/` builds the locked PyGObject, requiring glib 2.80+ and the
  headers above. On Debian 12 or Fedora 39 and older, use `./install.sh --dev`,
  then `venv/bin/pytest` or `venv/bin/python -m vocalinux.main --debug` instead.
- `venv/` uses distro PyGObject via `--system-site-packages`. The installer ignores
  an activated venv and rebuilds one created by the wrong interpreter. It uses
  `/usr/bin/python3` unless overridden, e.g. `SYSTEM_PYTHON=/usr/bin/python3.12 ./install.sh`.

## Making changes

### Branch naming

```bash
git checkout main
git pull origin main
git checkout -b feature/your-feature-name
# or fix/, docs/, refactor/, test/
```

Never push directly to `main`. Open a pull request for every change.

### Code style

| Tool | Role |
|------|------|
| Black | Formatting (line length 100) |
| isort | Import sorting (black profile) |
| flake8 | Linting |
| mypy | Type checking (`just typecheck`) |

```bash
just format    # black + isort
just lint      # flake8 + black --check + isort --check
just typecheck # mypy src/
```

### Project structure

```
voculinux/
├── src/vocalinux/            # Application
│   ├── main.py
│   ├── speech_recognition/   # Engines + command processor
│   ├── text_injection/       # X11 / Wayland injection
│   ├── ui/                   # GTK tray, settings, config
│   └── utils/
├── tests/
├── resources/                # Icons and sounds
├── docs/
├── install.d/                # Installer modules, including generated package data
├── packaging/                # Packaging inputs; Void xbps is the published channel
├── scripts/                  # Maintenance tools and distro package-map source
├── snap/                     # Snap recipe (not published for Voculinux)
└── web/                      # Marketing site (Next.js)
```

| Task | Start here |
|------|------------|
| Voice command | `src/vocalinux/speech_recognition/command_processor.py` |
| UI | `src/vocalinux/ui/` |
| Recognition engines | `src/vocalinux/speech_recognition/recognition_manager.py` |
| Text injection | `src/vocalinux/text_injection/text_injector.py` |
| Settings / config | `src/vocalinux/ui/config_manager.py`, `settings_dialog.py` |
| Installer package names | `scripts/distro-package-map.yaml` |

### Installer package map

`scripts/distro-package-map.yaml` is the source of truth for package names,
alternatives, and conditional package probes used by `install.sh`. Do not edit
the generated `install.d/package_map.sh` directly. After changing the YAML, run:

```bash
just distro-packages
just distro-packages-check
```

Commit both the YAML and generated shell module. Package-manager control flow
and installation policy remain handwritten in
`install.d/system_dependencies.sh`; user systems never parse the YAML. The
remote bootstrap's pre-checkout `git` installation is the deliberate exception
because the generated module is not available yet.

## Testing

```bash
just test
just test-cov
uv run --extra dev --extra vad --group lint pytest tests/test_command_processor.py
```

- Place tests in `tests/` as `test_*.py`
- Aim for solid coverage on new code (roughly 80%+)
- Use `pytest-mock` via the `mocker` fixture

```python
def test_command_processor_new_line(mocker):
    """Test that 'new line' command returns correct action."""
    processor = CommandProcessor()
    result = processor.process("new line")
    assert result.action == "new_line"
```

### Remote API test server

```bash
python scripts/test_remote_server.py
python scripts/test_remote_server.py --port 9000
python scripts/test_remote_server.py --delay 2
```

Endpoints:

- whisper.cpp style: `http://localhost:8080/inference`
- OpenAI-compatible: `http://localhost:8080/v1/audio/transcriptions`

In Settings → Speech Model, select **Remote API**, set the server URL, choose the endpoint format, and use **Test Connection**.

## Pull requests

> Automated agents may append three robot emoji (🤖) to the PR title to opt into the agent fast-track merge path when that process is active.

### Before you open a PR

- [ ] Style checks pass (`just lint`, `just typecheck`)
- [ ] Tests pass (`just test`)
- [ ] Docs updated when behavior or install steps change
- [ ] Commits use [Conventional Commits](https://www.conventionalcommits.org/)

```
type(scope): short description

Longer description if needed.

Fixes #123
```

Types: `feat`, `fix`, `docs`, `style`, `refactor`, `test`, `chore`

Examples:

```
feat(commands): add "select all" voice command
fix(tray): resolve icon not updating on Wayland
docs(readme): update installation instructions
```

### Submit

1. Push your branch to your fork
2. Open a PR against `main` and fill out the template
3. Link related issues
4. Wait for CI; address review feedback
5. Maintainers squash-merge when approved

## Releases

Maintainers follow [docs/RELEASE_PROCESS.md](docs/RELEASE_PROCESS.md) (version files, docs, website, tag, automated publish). Do not tag releases from feature branches.

## Community

- [SUPPORT.md](SUPPORT.md) - where to get help
- [Issues](https://git.bongbetic.com/xavierk/voculinux/issues)

## License

Contributions are licensed under the [GNU Affero General Public License v3.0](LICENSE)
(AGPL-3.0). By opening a pull request, you agree that your contribution may be
distributed under that license.

Voculinux is a hard fork of Vocalinux by VocaHQ: https://github.com/VocaHQ/vocalinux